Thank you for the clarification. It works!
Sathish,The constraint you described is Marathon's, not Mesos's :)Spark.mesos.constraints are applied to slave attributes like tachyon=true;us-east-1=false, as described in https://issues.apache.org/jira/browse/SPARK-6707.Cheers,-MaoOn Fri, Jan 29, 2016 at 2:51 PM, Sathish Kumaran Vairavelu <firstname.lastname@example.org> wrote:Hi
Quick question. How to pass constraint [["hostname", "CLUSTER", "specific.node.com"]] to mesos?
I was trying --conf spark.mesos.constraints=hostname:specific.node.com. But it didn't seems working
SathishOn Thu, Jan 28, 2016 at 6:52 PM Mao Geng <email@example.com> wrote:From my limited knowledge, only limited options such as network mode, volumes, portmaps can be passed through. See https://github.com/apache/spark/pull/3074/files.https://issues.apache.org/jira/browse/SPARK-8734 is open for exposing all docker options to spark.-MaoOn Thu, Jan 28, 2016 at 1:55 PM, Sathish Kumaran Vairavelu <firstname.lastname@example.org> wrote:Thank you., I figured it out. I have set executor memory to minimal and it works.,
Another issue has come.. I have to pass --add-host option while running containers in slave nodes.. Is there any option to pass docker run parameters from spark?On Thu, Jan 28, 2016 at 12:26 PM Mao Geng <email@example.com> wrote:Sathish,I guess the mesos resources are not enough to run your job. You might want to check the mesos log to figure out why.I tried to run the docker image with "--conf spark.mesos.coarse=false" and "true". Both are fine.Best,MaoOn Wed, Jan 27, 2016 at 5:00 PM, Sathish Kumaran Vairavelu <firstname.lastname@example.org> wrote:Hi,On the same Spark/Mesos/Docker setup, I am getting warning "Initial Job has not accepted any resources; check your cluster UI to ensure that workers are registered and have sufficient resources". I am running in coarse grained mode. Any pointers on how to fix this issue? Please help. I have updated both docker.properties and spark-default.conf with
spark.mesos.executor.docker.image and other properties.
SathishOn Wed, Jan 27, 2016 at 9:58 AM Sathish Kumaran Vairavelu <email@example.com> wrote:Thanks a lot for your info! I will try this today.On Wed, Jan 27, 2016 at 9:29 AM Mao Geng <firstname.lastname@example.org> wrote:Hi Sathish,The docker image is normal, no AWS profile included.When the driver container runs with --net=host, the driver host's AWS profile will take effect so that the driver can access the protected s3 files.Similarly, Mesos slaves also run Spark executor docker container in --net=host mode, so that the AWS profile of Mesos slaves will take effect.Hope it helps,Mao
On Jan 26, 2016, at 9:15 PM, Sathish Kumaran Vairavelu <email@example.com> wrote:Hi Mao,I want to check on accessing the S3 from Spark docker in Mesos. The EC2 instance that I am using has the AWS profile/IAM included. Should we build the docker image with any AWS profile settings or --net=host docker option takes care of it?Please helpThanksSathishOn Tue, Jan 26, 2016 at 9:04 PM Mao Geng <firstname.lastname@example.org> wrote:Thank you very much, Jerry!I changed to "--jars /opt/spark/lib/hadoop-aws-2.7.1.jar,/opt/spark/lib/aws-java-sdk-1.7.4.jar" then it worked like a charm!From Mesos task logs below, I saw Mesos executor downloaded the jars from the driver, which is a bit unnecessary (as the docker image already has them), but that's ok - I am happy seeing Spark + Mesos + Docker + S3 worked together!Thanks,Mao16/01/27 02:54:45 INFO Executor: Using REPL class URI: http://172.16.3.98:33771 16/01/27 02:55:12 INFO CoarseGrainedExecutorBackend: Got assigned task 0 16/01/27 02:55:12 INFO Executor: Running task 0.0 in stage 0.0 (TID 0) 16/01/27 02:55:12 INFO Executor: Fetching http://172.16.3.98:3850/jars/hadoop-aws-2.7.1.jar with timestamp 1453863280432 16/01/27 02:55:12 INFO Utils: Fetching http://172.16.3.98:3850/jars/hadoop-aws-2.7.1.jar to /tmp/spark-7b8e1681-8a62-4f1d-9e11-fdf8062b1b08/fetchFileTemp1518118694295619525.tmp 16/01/27 02:55:12 INFO Utils: Copying /tmp/spark-7b8e1681-8a62-4f1d-9e11-fdf8062b1b08/-19880839621453863280432_cache to /./hadoop-aws-2.7.1.jar 16/01/27 02:55:12 INFO Executor: Adding file:/./hadoop-aws-2.7.1.jar to class loader 16/01/27 02:55:12 INFO Executor: Fetching http://172.16.3.98:3850/jars/aws-java-sdk-1.7.4.jar with timestamp 1453863280472 16/01/27 02:55:12 INFO Utils: Fetching http://172.16.3.98:3850/jars/aws-java-sdk-1.7.4.jar to /tmp/spark-7b8e1681-8a62-4f1d-9e11-fdf8062b1b08/fetchFileTemp8868621397726761921.tmp 16/01/27 02:55:12 INFO Utils: Copying /tmp/spark-7b8e1681-8a62-4f1d-9e11-fdf8062b1b08/8167072821453863280472_cache to /./aws-java-sdk-1.7.4.jar 16/01/27 02:55:12 INFO Executor: Adding file:/./aws-java-sdk-1.7.4.jar to class loaderOn Tue, Jan 26, 2016 at 5:40 PM, Jerry Lam <email@example.com> wrote:Hi Mao,Can you try --jars to include those jars?Best Regards,Jerry
Sent from my iPhone
On 26 Jan, 2016, at 7:02 pm, Mao Geng <firstname.lastname@example.org> wrote:Hi there,I am trying to run Spark on Mesos using a Docker image as executor, as mentioned http://spark.apache.org/docs/latest/running-on-mesos.html#mesos-docker-support.I built a docker image using the following Dockerfile (which is based on https://github.com/apache/spark/blob/master/docker/spark-mesos/Dockerfile):FROM mesosphere/mesos:0.25.0-0.2.70.ubuntu1404# Update the base ubuntu image with dependencies needed for SparkRUN apt-get update && \apt-get install -y python libnss3 openjdk-7-jre-headless curlRUN curl http://www.carfab.com/apachesoftware/spark/spark-1.6.0/spark-1.6.0-bin-hadoop2.6.tgz | tar -xzC /opt && \ln -s /opt/spark-1.6.0-bin-hadoop2.6 /opt/sparkENV SPARK_HOME /opt/sparkENV MESOS_NATIVE_JAVA_LIBRARY /usr/local/lib/libmesos.soThen I successfully ran spark-shell via this docker command:docker run --rm -it --net=host <registry>/<image>:<tag> /opt/spark/bin/spark-shell --master mesos://<master_host>:5050 --conf <registry>/<image>:<tag>So far so good. Then I wanted to call sc.textFile to load a file from S3, but I was blocked by some issues which I couldn't figure out. I've read https://dzone.com/articles/uniting-spark-parquet-and-s3-as-an-alternative-to and http://blog.encomiabile.it/2015/10/29/apache-spark-amazon-s3-and-apache-mesos, learned that I need to add hadood-aws-2.7.1 and aws-java-sdk-2.7.4 into the executor and driver's classpaths, in order to access s3 files.So, I added following lines into Dockerfile and build a new image.RUN curl https://repo1.maven.org/maven2/com/amazonaws/aws-java-sdk/1.7.4/aws-java-sdk-1.7.4.jar -o /opt/spark/lib/aws-java-sdk-1.7.4.jarRUN curl http://central.maven.org/maven2/org/apache/hadoop/hadoop-aws/2.7.1/hadoop-aws-2.7.1.jar -o /opt/spark/lib/hadoop-aws-2.7.1.jarThen I started spark-shell again with below command:docker run --rm -it --net=host <registry>/<image>:<tag> /opt/spark/bin/spark-shell --master mesos://<master_host>:5050 --conf <registry>/<image>:<tag> --conf spark.executor.extraClassPath=/opt/spark/lib/hadoop-aws-2.7.1.jar:/opt/spark/lib/aws-java-sdk-1.7.4.jar --conf spark.driver.extraClassPath=/opt/spark/lib/hadoop-aws-2.7.1.jar:/opt/spark/lib/aws-java-sdk-1.7.4.jarBut below command failed when I ran it in spark-shell:scala> sc.textFile("s3a://<bucket_name>/<file_name>").count()[Stage 0:> (0 + 2) / 2]16/01/26 23:05:23 WARN TaskSetManager: Lost task 0.0 in stage 0.0 (TID 0, ip-172-16-14-203.us-west-2.compute.internal): java.lang.RuntimeException: java.lang.ClassNotFoundException: Class org.apache.hadoop.fs.s3a.S3AFileSystem not foundat org.apache.hadoop.conf.Configuration.getClass(Configuration.java:2074)at org.apache.hadoop.fs.FileSystem.getFileSystemClass(FileSystem.java:2578)at org.apache.hadoop.fs.FileSystem.createFileSystem(FileSystem.java:2591)at org.apache.hadoop.fs.FileSystem.access$200(FileSystem.java:91)at org.apache.hadoop.fs.FileSystem$Cache.getInternal(FileSystem.java:2630)at org.apache.hadoop.fs.FileSystem$Cache.get(FileSystem.java:2612)at org.apache.hadoop.fs.FileSystem.get(FileSystem.java:370)at org.apache.hadoop.fs.Path.getFileSystem(Path.java:296)at org.apache.hadoop.mapred.LineRecordReader.<init>(LineRecordReader.java:107)at org.apache.hadoop.mapred.TextInputFormat.getRecordReader(TextInputFormat.java:67)at org.apache.spark.rdd.HadoopRDD$$anon$1.<init>(HadoopRDD.scala:237)at org.apache.spark.rdd.HadoopRDD.compute(HadoopRDD.scala:208)at org.apache.spark.rdd.HadoopRDD.compute(HadoopRDD.scala:101)at org.apache.spark.rdd.RDD.computeOrReadCheckpoint(RDD.scala:306)at org.apache.spark.rdd.RDD.iterator(RDD.scala:270)at org.apache.spark.rdd.MapPartitionsRDD.compute(MapPartitionsRDD.scala:38)at org.apache.spark.rdd.RDD.computeOrReadCheckpoint(RDD.scala:306)at org.apache.spark.rdd.RDD.iterator(RDD.scala:270)at org.apache.spark.scheduler.ResultTask.runTask(ResultTask.scala:66)at org.apache.spark.scheduler.Task.run(Task.scala:89)at org.apache.spark.executor.Executor$TaskRunner.run(Executor.scala:213)at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1145)at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:615)at java.lang.Thread.run(Thread.java:745)Caused by: java.lang.ClassNotFoundException: Class org.apache.hadoop.fs.s3a.S3AFileSystem not foundat org.apache.hadoop.conf.Configuration.getClassByName(Configuration.java:1980)at org.apache.hadoop.conf.Configuration.getClass(Configuration.java:2072)... 23 moreI checked hadoop-aws-2.7.1.jar, the org.apache.hadoop.fs.s3a.S3AFileSystem class file is in it. I also checked the Environment page of driver's Web UI at 4040 port, both hadoop-aws-2.7.1.jar and aws-java-sdk-1.7.4.jar are in the Classpath Entries (system path). And following code ran fine in spark-shell:scala> val clazz = Class.forName("org.apache.hadoop.fs.s3a.S3AFileSystem")clazz: Class[_] = class org.apache.hadoop.fs.s3a.S3AFileSystemscala> clazz.getClassLoader()res2: ClassLoader = sun.misc.Launcher$AppClassLoader@770848b9So, I am confused why the task failed with "java.lang.ClassNotFoundException" Exception? Is there something wrong in the command line options I used to start spark-shell, or in the docker image, or in the "s3a://" url? Or is something related to the Docker executor of Mesos? I studied a bit https://github.com/apache/spark/blob/branch-1.6/core/src/main/scala/org/apache/spark/scheduler/cluster/mesos/MesosSchedulerBackend.scala but didn't understand it well...Appreciate if anyone will shed some lights on me.Thanks,Mao Geng